the ”ugric-turkic battle”: a critical review … · the ”ugric-turkic battle”: a critical...

ANGELA MARCANTONIO (Roma), PIRJO NUMMENAHO (Napoli),MICHELA SALVAGNI (Roma)

THE ”UGRIC-TURKIC BATTLE”: A CRITICAL REVIEW

1. Introduction

In Uralic studies it is usually taken for granted that the uniqueness of theFinno-Ugric node, and therefore the existence of a unique Finno-Ugric fam-ily, was established in the final decades of the 19th century using system-atic, scientific methods of analysis. This is based on supposedly compellinglinguistic evidence, mainly from J. Budenz, a German linguist active inHungary in the last decades of the 19th century.

J. Budenz is generally believed to have been the scientist who establishedthe existence and uniqueness of the Finno-Ugric node and family beyond doubt.Much of the Uralic literature is based on this belief. However, there appearsto be little discussion in the literature actually referring to the original worksof J. Budenz. The purpose of this paper is to review this original work critically.

As we shall see, there are two problems with the evidence put forwardby J. Budenz. Firstly, although J. Budenz claims to have adopted a systematicmethod of analysis (based on the Comparative Method), in fact he does nosuch thing. For example, he does not state the phonological criteria that areadopted for establishing the correspondences, so that it is not possible to ascer-tain their validity. In fact, 81% of a significant sample of his correspondencesare no longer considered valid in the modern literature. Secondly, even ifJ. Budenz’ conclusions had been supported by a testable method of systematicanalysis, the reported evidence is in fact at variance with the modern assump-tion that the Finno-Ugric node is unique. In fact, J. Budenz recognises a signifi-cant number of Turkic/Hungarian correspondences, and concludes that someof these are indicative of a genetic relationship between Hungarian and Turkic!

1.1. Overview

It was known at the time of J. Budenz that the Mediaeval Chronicles1 hadindicated an unspecified Eastern homeland for the Hungarians. Between

81

1 The main Mediaeval Chronicles that deal with Hungarian pre-history and historyare: ”Anonymus Gesta Hungarorum”, written circa 1200 AD, and ”Gesta Hun-garorum”, written by Simon Kézai between 1282—1285.

1 Linguistica Uralica 2 2001

the 15th and 17th centuries it came to be taken for granted that this home-land could be identified with an area near the Ural mountains called Yugria(hence the terms ”Uralic” and ”Ugric”). This belief was based on the appar-ent similarity between the toponym Yugria and the ethnonym hungarus.This connection was supported by a later discovery that one of the pop-ulations in the area (the Voguls) called themselves Mansi, which, accord-ing to B. Kálmán ”to the lay ear slightly resembles the name magyar” (1988: 395).

The connection magyar/Mansi is still regarded today as strong evi-dence in support of the Finno-Ugric/Uralic theory, and it is still assumedto this day that Hungarian shares a privileged relationship with the lan-guages in the Yugria area. Vogul, Ostyak and Hungarian are still held toform the conventional Ugric node, even though it is now generally recog-nised that Hungarian is radically different in phonology, morphology, lex-icon and syntax from the other Ugric languages (Abondolo 1987; Sam-mallahti 1988; Helimski 1984 : 253).

In a period of about thirty years to the end of the 19th century researchersbegan to apply Darwinian models to language development and to use thenewly emerged Comparative Method. In the meantime, the Ugric theory,that is, the belief that Hungarian, Vogul and Ostyak are genetically related,was extended to include the languages of northern Europe (such as Finnish,Lapp, Mordvin, Zyrian etc.), and put on a supposedly scientific footing. Infact, J. Budenz first attempted to apply the Comparative Method to examinethe correlations known at the time to exist between Hungarian and Turkicon the one hand, and between Hungarian and the ”Ugric” languages (Vogul,Ostyak, Finnish, Lapp, etc.) on the other. Starting in the 1870s, J. Budenz,through a number of publications, argued that Hungarian was more closelyrelated to the Ugric languages than it was to Turkic. Therefore, it was tobe classified as Ugric. He supported this conclusion with a corpus of lex-ical Hungarian/Ugric correspondences.

However, as we shall see, in these original works the ComparativeMethod was either not applied correctly, or was not applied at all, for twobasic reasons. Firstly, the Comparative Method was in its infancy at thetime, which means that it could not have been applied systematically andscientifically in any case. In fact, this holds true not only of the Uralic lan-guages, but also of the Indo-European languages, for the examination ofwhich the method emerged and developed. As A. Fox says, ”the forms ofreconstructed Proto-Indo-European have changed out of all recognition assuccessive generations of scholars have refined and amended their pre-decessors’ work” (1995 : 11).

Secondly, despite his correct programmatic and methodological inten-tions, J. Budenz did not specify the sound-rules by which he tried to estab-lish his Hungarian/Ugric correspondences. Furthermore, arbitrary stretchesof meaning were often accepted in order to establish the desired corre-spondence between Hungarian and a Ugric language, or to reject an unde-sirable Hungarian/Turkic correspondence. In a word, Budenz’ corpusturns out to be of very poor quality, particularly in the light of modernlinguistic knowledge.

In addition to this, it usually goes unmentioned in the literature thatJ. Budenz’ conception of the (Finno-)Ugric family differs significantly from

Angela Marcantonio, Pirjo Nummenaho, Michela Salvagni

82

the modern theory he is supposed to have scientifically established. At theend of the 19th century it was universally accepted among scholars thatthere was an overarching genetic relationship spanning all the languagesthat are now classified as Uralic and Altaic. J. Budenz, and the other scholarswho took part in the ’Ugric-Turkic battle’, embraced this view whole-heartedly. In particular, J. Budenz classified Hungarian as belonging to theextended Ugric branch, and, in his model, the Ugric branch, the Turkicbranch and other Asiatic branches, all belonged to the wide Altaic family.As J. Budenz says (1869 : 375): ”A nagy altaji [sic] nyelvcsaládot egyesnyelvcsoportok teszik: 1) mongol-mandsu-török-±; 2) finn-ugor-±; 3) szamojed[sic], melyek a szóalakulás jellemére, a származott szók és viszonyított nyel-valakok miképen való képzésére, s a szók összeszerkesztésének néhányf é Óovonásaira nézve egymással egyezé Óoknek mutatkoznak.2

J. Budenz even supported this model by identifying specific corre-spondences that he claimed were indicative of a genetic relationship (froma higher level node) between Turkic and Hungarian.

In contrast to this picture, in the modern theory the Uralic languagesare held to be completely unrelated to the Altaic languages. This is becauseat some point in the development of the paradigm (between the end of the19th century and the beginning of the 20th century), the Uralic languagescame to be split off from the languages left behind in the Altaic family,although no original work to substantiate this assumed splitting-off can befound (Georg, Michalove, Manaster Ramer, Sidwell 1998). One consequenceof this is that all the words of Turkic origin present in Hungarian are nowclassified as loan-words. Yet, J. Budenz made no such claim. J. Budenzsimply intended to prove that, within the context of the wide Altaic fam-ily, Hungarian shares many more features with the Ugric languages thanwith the Turkic languages, and it is therefore to be definitively classifiedas Ugric.

Last but not least, it is worth observing that even the claim that Hungar-ian is more closely related to the Ugric languages than to Turkic turns outto be incorrect if one expects this ’closeness’ to be substantiated by a sig-nificant number of shared features. In fact, as we shall see, within a givensample of J. Budenz’ correspondences, the number of correspondencesHungarian shares with the Altaic languages (and Yukaghir) in fact nearlymatch the number of correspondences Hungarian shares with the (Finno±-±)Ugric/Uralic languages, according to modern knowledge.

As is evident from what said above, the term ”Ugric” in J. Budenz’works does not have the same meaning as within the modern, conven-tional paradigm. In fact, J. Budenz did not even attempt to reconstructeither the conventional Ugric node (Hungarian, Vogul and Ostyak), or theconventional Finno-Ugric node, or those other intermediate nodes whichrepresent a cornerstone of the modern family tree (notice however that theauthor occasionally mentions the definition ”Finno-Ugric”, for example inhis article of 1869; see quote above). In other words, by using the term

The ”Ugric-Turkic Battle „”: A Critical Review

83

2 ’The wide Altaic language group is formed by various individual groups: 1) Mon-golian-Manchu-Turkic; 2) Finno-Ugric; 3) Samoyed. These groups, judging fromthe nature of the development of the words, the formation of the derived wordsand the compared language forms, as well as some general tendencies of assem-bling the words, show similarities with each other’ (our translation).

1*

”Ugric” J. Budenz (1871—1873; 1878/1879) simply meant that the Ugricbranch (of the Altaic family) was not divided into sub-groups or sub-nodes,but directly into main languages (Hauptsprachen). More precisely, in Bu-denz 1878/1879 : 196 the Ugric languages were divided into two majorgroups: a) North-Ugric, which includes: Lapp, Permian (Zyrian and Votyak),Ob-Ugric (Vogul and Ostyak) and Hungarian; b) South-Ugric, which includesBalto-Finnic, Mordvin and Cheremis. This division is motivated by the dis-tribution of initial Én- (North-Ugric) vs n- (South-Ugric); for more detail onthis topic and on other types of Uralic family tree see Sutrop 2000.

2. The ”Ugric-Turkic battle”

2.1. The presumed results

It is commonly claimed that the existence of the Finno-Ugric family (notthe Uralic family as a whole yet) was first proven beyond doubt by a groupof scholars, directed by the leading figure of József Budenz, as the resultof an intense academic debate known as Az ugor-török hÉaborÉu (the Ugric-Turkic battle). The existence of the Finno-Ugric family is held to have sub-sequently been reinforced by the work of Otto Donner on the close andunique relationship between the Samoyed languages and the newly estab-lished Finno-Ugric family, thus giving rise to the Uralic family as it is tra-ditionally conceived of. In fact, the conventional Uralic family tree datesback to O. Donner’s family tree outlined in his paper of 1879/1880.

One also often reads that, as a consequence of the outcome of the Ugric-Turkic battle, the competing thesis of the Turkic origin of Hungarian wasdefinitively defeated. As M. Ruhlen (1987 : 68) says: ”In the final decadesof the nineteenth century.... Budenz, [spent] the last 20 years of his life asthe leading Finno-Ugric specialist in Hungary. In his major work, a com-parative dictionary of the Finno-Ugric elements in Hungarian, Budenz pro-posed close to 1,000 correspondences between Hungarian and variousFinno-Ugric languages, and laid to rest forever the erroneous notions con-cerning Hungarians closest relatives.”

Similarly A. Róna-Tas3 (1978 : 263—264) says ”In 1869 Á. Vámbéry pub-lished a paper on the Hungarian and ”Turkish-Tatar” word comparisons. Themajor but methodologically insufficiently gathered material was severelycriticised by J. Budenz in 1871, and with this the so-called ”Turkish-Ugrian4

battle” began. ”Ugrian” in those times was used for what we now callFinno-Ugrian. On the one side stood Vámbéry and his followers who triedto defend the thesis that the Hungarians and their language are of Turk-ish origin.... On the other hand J. Budenz and other scholars offered sci-entific arguments in favour of the Finno-Ugrian origin of the Hungarianlanguage. They insisted that the Turkish elements in the Hungarian lan-guage are loan-words from several periods.... Many Hungarians rejectedthe relationship with poor people with a ”fish fat smell” and enthusiasti-


84

3 The quote is from footnote (7), starting at p. 263.4 Note that the author uses the term ”Turkish”, which is in fact the correct Eng-lish translation of the Hungarian word. However, given that the debate revolvesaround the correlations between Hungarian and the various Turkic languages, andnot just Turkish, it is better to use the term ”Turkic”, as we do here.

cally ”fought” for the ”glorious Turkish origin”. It was to the merit of Bu-denz, Hunfalvy, Munkácsi, Szinnyei and others that the up-to-date methodsand sound facts slowly overcame the sentiments.”

2.2. The real results

The real results of the Ugric-Turkic battle are quite different from what isgenerally propagated and believed. As we shall see, it is not true that J. Bu-denz considered a l l of Á. Vámbéry’s Hungarian/Turkic correspondencesas incorrect. It is not true either that J. Budenz ”insisted that the Turkishelements in the Hungarian language are loan-words from several periods”,or that the works of J. Budenz and colleagues are based on ”up-to-datemethods and sound facts”. Quite the contrary. The true picture is sum-marised in the following three points:• J. Budenz recognised the existence of a certain number of correct cor-respondences between Turkic and Hungarian, among the many incorrectparallels proposed by Á. Vámbéry;• most of these correct correspondences were considered by the author tobe indicative of a remote, genetic relationship between Hungarian and Tur-kic (that this was the real situation is occasionally acknowledged in theliterature, see for example Räsänen 1963/1964 : 184 and Pusztay 1977 : 95);• as mentioned, the Comparative Method was not — and could not, infact — be applied properly.

Recall at this point that J. Budenz does not take into account the Samoyedlanguages. In other words, he establishes correspondences only between Hun-garian and one or other of the Finnic and/or Ugric languages. Notice alsothat we examine only the works by J. Budenz, and not the works by theother authors who also took part to the debate, for which see Pusztay 19775.

3. Budenz corpus

In this paragraph we shall examine in detail a significant sample of J. Bu-denz’ Hungarian/Ugric correspondences. These correspondences, which,according to the author, should prove beyond doubt the belonging-nessof Hungarian to the Ugric family, will in fact support the picture outlinedin the three points above.

For the purpose of our analysis we have gone back to the original pa-per by Á. Vámbéry of 1869 and to the original paper by Budenz of 1871—1873, as well as to J. Budenz’ dictionary: ”Magyar-ugor összehasonléıt Éosz Éot Éar” (MUSz; 1873—1881). J. Budenz (1871—1873) argues that most, butnot all, of the Hungarian/Turkic parallels proposed by his adversary arewrong. The author divides them into 4 classes:• Class I contains correct correspondences; these can be considered testi-mony either of an original, genetic Ugric/Turkic relationship (34 words areclassified this way), or of borrowing (8 words are classified this way). Toquote his own words (1871—1873 : 75) ”a szóegyezéseket négy csoportraosztva sorolom elé: I. Olyanokat, melyek helyeseknek fogadhatók el, vagy


85

5 Among the other authors who took part to the debate see for example P. Hun-falvy (1883). See also the dictionary by J. Szinnyei ”Magyar nyelv hasonlítás” (1896—1927).

mint a törökb éÓol való magyar kölcsönvételek, vagy mint az ”ugor-török éÓos-rokonság” tárgyalásához tartozó adatok.”6

• Class II contains correspondences which appear to be good at first sight,but are in reality incorrect (lÉatszatos egyezÉesek ’apparent correspondences’).• Class III contains correspondences which are clearly wrong (helytelenegyezÉesek ’wrong correspondences’).• Class IV contains irrelevant correspondences in the sense that they are ei-ther clear recent borrowing, or onomatopoeic, or irrelevant for other reason.

For each of the incorrect Hungarian/Turkic parallels proposed by Á. Vám-béry and then listed in Classes II and III, J. Budenz proposes his (alter-native) correspondences from one or the other Finnic or Ugric languages.He thus intended to prove, on the basis of phonological and/or semanticreasons, that the Hungarian/Finnic/Ugric parallels, and not the Hungar-ian/Turkic parallels, are correct.

We have analysed a sample of Budenz Corpus. We have selected a l lthe etymologies which can be classified as belonging to the core-Lexicon(which tends to be more resistant to the process of borrowing and morestable with regard to phonetic and semantic changes). More precisely, wehave examined all the (core-Lexicon) items contained in Á. Vámbéry’s paper7

of 1869, which have been subsequently reported and analysed by J. Bu-denz (first in his paper of 1871—1873 and then in his dictionary). We haveanalysed only the core-Lexicon parallels included in Classes I—III, whichnumber a total of 123 items — the etymologies of Class IV have not beentaken into consideration, being indeed irrelevant. We have examined theparallels of Class I, that is the ’correct’ parallels, at first separately fromthe ’incorrect’ parallels (included in Classes II and III), and afterwards to-gether with them. We have also compared all of these 123 items againstthe modern ”Uralisches Etymologisches Wörterbuch” (UEW), to verify whattheir status is in the light of modern knowledge.8


86

6 ’I will list the word correspondences dividing them into four groups: group I:those which are to be considered correct, either as loan-words from Turkic, or asdata pertaining to the discussion of the ”Ugric-Turkic original genetic relationship” ’(our translation).7 Á. Vámbéry (1869) stated that Hungarian belongs to the (Finno-)Ugric languagegroup and considered that the Turkic elements present in it are the effect of a rela-tionship he classified as of second nature/degree. This relationship is in fact dueto close, recent, intense contacts, basically to összeolvadÉas ’fusion’. However, in 1882Á. Vámbéry published a book called: ”A magyarok eredete; ethnologiai tanulmÉany”,where he strongly claimed that the Hungarian people and language are of Turkicorigin. He based his claim mainly on the following two elements: a) the historicalsources, which consistently call the Hungarian Turks; b) the many lexical and struc-tural similarities between the two languages.8 It is acknowledged that the most comprehensive source for the Uralic languagesis K. Rédei’s UEW. Therefore, we have used this dictionary (the only Uralic dic-tionary available) as a reference point against which to compare Budenz corpus.we have also used UEW transcription (unless otherwise specified) except, of course,when examples are reported within a quote. The purpose of this is for the readerto retrieve the reference and compare the original languages, and it does not neces-sarily denote general acceptance of UEW reconstruction in the literature or by theauthors of this paper. For a review of the general acceptance or otherwise of UEWreconstructions see Kulonen 1996. See also Marcantonio 2001, where it is arguedthat many of the Uralic etymologies listed in UEW could, in fact, also be the resultof chance resemblance.

3.1. Class I: the ”correct correspondences”

The first relevant factor which emerges from the analysis of the Hungarian/Turkic parallels classified by J. Budenz as belonging to Class I is that theyturn out to be quite a good number. In fact, of the total number of 123 par-allels examined, the correct correspondences number 42 items. Furthermore,the majority of these 42 good correspondences, exactly 34, are considered byJ. Budenz as indicative of old, original genetic relationship between Turkicand Hungarian, whilst only 8 are considered to be the result of borrowing.This is, of course, a significant result, and one that is hardly ever reportedin the literature. Notice, furthermore, that there does not seem to be anycriteria (or at least, they are not specified), on the basis of which J. Budenzmakes his judgement as to whether these correspondences are possible cog-nates or possible loan-words. J. Budenz himself seems to recognise this factby marking with a question mark the one or the other sub-classification.Notice also that for most of these good Hungarian/Turkic parallels J. Bu-denz also reports correspondences from the Ugric languages.

In Table I at the end of the paper we have reported the 42 Hungarianwords included in Class I, without, however, reporting the Turkic paral-lels as proposed by Á. Vámbéry (this being not relevant for the aim ofthis paper). For each of the listed words we have indicated the page num-ber in J. Budenz’ MUSz dictionary or, alternatively, in J. Budenz’ paper of1871—1873 (indicated as NyK), if the word in question is not reported inthe dictionary. We have also indicated the classification of these wordsaccording to UEW, that is, we have indicated to which Uralic branch theword belongs and whether there are correspondences in non-Uralic lan-guages. If the word under discussion is not recognised as being of Finno-Ugric or Uralic origin — and therefore is not reported — by UEW, the abbre-viation n.l. ’not-listed’ is used. If the Hungarian word (and related Uralicetymology) also has a non-Uralic parallel this is indicated by adding thesymbol +. Table I has been divided into two sections: the first section,Table Ia, contains those 34 Hungarian parallels (of the Turkic words) which,according to J. Budenz, are indicative of an original relationship (é Óosrokon-sÉag); the second section, Table Ib, contains those 8 parallels believed by J. Bu-denz to be the result of borrowing.

The following comments may help evaluate the content of Table I.Regarding Table Ia the following holds:• 11 Hungarian words are not listed in UEW, that is, they are not recog-nised as being of Uralic origin by the modern dictionary (items number4, 7, 10, 11, 12, 13, 18, 24, 31, 33, 34); • 11 words are recognised as being of Uralic origin, but they also have aparallel in the Altaic languages and/or Yukaghir according to UEW (itemsnumber 1, 2, 8, 17, 20, 21, 22, 23, 28, 29, 30); • 9 words are considered to be of Uralic origin only, that is, no parallelsin non-Uralic languages are reported.

Regarding Table Ib, still according to UEW, the following holds:• 2 Hungarian words are not listed (items number 2 and 8); • 2 words are connected with Altaic languages, although with uncertainty(number 1 and 3); • 1 word has good Uralic as well as Turkic correspondences (item number 6);


87

• 1 word is classified as an uncertain Hungarian parallel of an otherwisegood Uralic etymology; a derivation from Turkic is considered possible(item number 4); • 2 words are classified as having good correspondences in Uralic lan-guages only (items number 5 and 7).

3.2. Class II and III: the ”incorrect correspondences”

3.2.1. The list of the incorrect correspondences

Let us now discuss the incorrect correspondences as listed by J. Budenzin Classes II and III. They can be analysed together, despite the fact thatJ. Budenz divides them into two different groups, given that the differ-ence between them is fictitious. The parallels in question, let us recall, areHungarian words which, according to Á. Vámbéry, have good corre-spondences in the Turkic languages; according to J. Budenz, however, theproposed Turkic/Hungarian correspondences are incorrect, for phonolog-ical or semantic reasons, or both. Instead, J. Budenz proposes alternativeHungarian/Ugric parallels which he claims to be correct. These alterna-tives, supposedly good Hungarian/Ugric parallels — 81 in total — havebeen compared against the UEW dictionary, in the same way as the cor-respondences of Class I have, to verify their status in the light of modernknowledge. The results of this comparison are as follows:• only 23 Hungarian words have good, purely Finno-Ugric/Uralic corre-spondences; that is, out of the 81 Hungarian words classified by J. Budenzas Ugric UEW establishes correspondences solely within the Uralic area foronly 23 of them (the full list of these etymologies is reported in Table II atthe end of the paper);• 13 words have good Finno-Ugric/Uralic counterparts, but UEW also reportsgood correspondences with non-Uralic languages, mainly the Altaic lan-guages, but also Yukaghir or, more rarely, some Indo-European languages;• 11 words are considered to be part of an uncertain Finno-Ugric/Uralicetymological set (the etymologies are written in italics, which, accordingto UEW, indicates a high degree of uncertainty);• in 3 cases, although the overall Uralic etymology is considered valid,the Hungarian parallel itself is considered uncertain;• 30 Hungarian words are not listed in UEW at all;• there is one duplicate.

3.2.2. The quality of the Ugric corpus

It is often the case that J. Budenz’ choice of one or other of the Finnic orUgric parallels is wrong, even if the overall Hungarian/Ugric connectionis correct. For example, J. Budenz connects the Hungarian word for ’hunger,to be hungry’, ÉehsÉeg, Éehes (MUSz 771) with Finn. hiuka (Gen. hiua-n, hiukais-ta ’to feel hungry’), and Cher. çsuçz- ’to be hungry’. According to UEW 208and 851, the Hungarian words are not at all connected to the Finnish andCheremis equivalents, which are, in fact, not listed.

Similarly, J. Budenz connects Hun. boka ’ankle, heels’ to Finn. pahka’gnarl, protuberance, node’, and Mord. pov, povka ’button, knob’ (MUSz


88

460). However, Hun. boka is not listed in UEW, and is considered of oldTurkic origin in TESz I 327. On the other hand, Finn. pahka, according toUEW 350, is connected to MordM pakçs (< P-U *pakçsa). In J. Budenz’ opin-ion Hun. orr ’nose’ is connected to Finn. turpa ’muzzle’, Mord. turva ’lip’,Zyr. t˙ırp, Ost. tor˙ıp (MUSz 842). According to UEW 801, the Finnish, Mord-vin and Zyrian correspondences are valid, but the Hungarian and Ostyakparallels are wrong, the etymology being classified only as Finno-Permian.Hungarian orr might instead be connected to the F-U root *w≠ere (*wºore)’mountain’, from which Zyr. v≠er and Vog. wºar- ’forest’ are derived (UEW571), but with uncertainty, due to the switch of meaning.

In addition to this, in our opinion there is another, a more serious prob-lem. Often the sound-shape and especially the meaning of a given set ofparallels are stretched beyond any reasonable interpretation in order totry and establish a connection between Hungarian and another Ugric lan-guage. Conversely, Hungarian and Turkic parallels with basically the samemeaning and similar sound-shape may be considered incorrect on the basisof difference in meaning (for example, ’hair (of body)’ vs ’hair (of head)’).A few paradigmatic examples will suffice. In example 1 below J. Budenzrejects the connection between similar words in Hungarian and Turkic for’navel’ on the grounds that the semantic match is not present in all theTurkic dialects. Instead he makes an improbable connection between theHungarian term for ’navel’ with the Ugric term for ’tongue’, ’dumb’ or’rope’, despite the fact that Hungarian has its own term for ’tongue’: nyelv,as recognised by the author himself.1a. According to Á. Vámbéry (1869 : 161), Hun. köldök ~ kõdök 9 ’navel’corresponds to Cha. köndük / Osm. köbek, göbek ’navel’.1b. According to J. Budenz (1871—1873 : 104), köldök belongs to Class II,that is, the Hungarian/Turkic connection is wrong. This is because Osm.göbek does not agree with the Hungarian form. The form containing -l-,in J. Budenz’ opinion, is Ugric — no explanation is given as to why thisshould be the case.1c. According to J. Budenz (MUSz 41—42), köldök is to be connected withLapp kiäl-dak ’chorda, fides, nervus’ / Finn. kieli ’tongue, language’ / Mord.käl ’tongue, language’ / Vog. kalÍ ’dumb, mute, speechless’, etc. This isbecause the real meaning of the Finnish word is held to be ’thin danglingbody’ (’fityeg é Óo vékony test’), as testified by the Finnish phrase kello-n kieli’bell-of tongue, tongue of bell’. Similarly, the Hungarian word would notreally mean ’navel’, but ’navel cord’ (’nabels schnur’).1d. köldök is not listed in UEW.

We are sure today that the Finnish term kieli ’tongue’ derives from atotally different root, reconstructed as Uralic *kele (*kºele)10 by UEW 144. Ithas parallels in most Uralic languages, as well as in the Altaic languagesand Yukaghir. It is also generally recognised that köldök is indeed connectedto similar terms present in Mongolian, Turkic and Samoyed, and that theHungarian term could have been borrowed directly from Turk. kin-dik (Ró-na-Tas 1988 : 744; Ligeti 1986 : 30, 81, etc.). Notice that J. Budenz is awareof the existence of words similar in sound and meaning present in the


89

9 kõdök = kéÓodök. 10 The Finnish dictionary SSA (1 1992 : 353) mentions the similarity between the Uralicand the Altaic forms, without taking position about the status of the similarity.

Turkic languages; for example, he mentions Tat. kündük, kindik; nevertheless,he simply states that one cannot doubt the Ugric nature of the word.

In example 2 J. Budenz rejects Á. Vámbéry’s connection between the Hun-garian and Turkic terms for ’to lay (eggs)’ on the grounds that the Hungar-ian term is a transitive verb, whilst the Turkic verb is intransitive. Instead,he gives a forced interpretation of Est. too- ’to bring’ as if it meant ’to layeggs’, on the grounds that this verb can co-occur with the noun muna ’egg’.2a. According to Á. Vámbéry (1869 : 180) Hun. tojni ’to lay eggs’, tojÉas’egg’, etc. are connected to Cha. togmak ’to be born’, togum ~ to˛um ’egg,seed, (child)-birth’, etc.2b. According to J. Budenz (MUSz 222—223) Hun. toj-ni and tojÉas (whichbelong to Class II) are to be connected to Est. too- ’to go and get, to bring’,because this verb occurs in the phrase muna too- ’to lay an egg’. Withinthe same etymological set are also included Finn. tuo-da and Mord. tuje-’to bring, get’. He argues that the Hungarian verbs toj- ’to lay eggs’ andhoz ’to bring, carry’ derive from the same Ugric form meaning ’to bring,carry’. The justification for the switch in meaning (from the more general’to bring’ to the more restricted ’to lay eggs’) is to be found in the simi-lar semantic switch that occurred in going from I-E *bhar- and Lat. fer-to Ger. gebären ’to give birth’.2c. According to UEW 528, the Hungarian verbs toj- and hoz are not con-nected. Finn. tuo- and Mord. tuje- can be derived from the reconstructed P-Uroot *to¸e- ’to bring, get, give’, to which toj- could be connected, but withuncertainty for several reasons, including the semantic switch. Comparealso Yuk. tadi ’to give’ / Old Ind. da-dºati ’he gives’ / Lat. dare (UEW 529).

In example 3 J. Budenz rejects Á. Vámbéry’s connection between theHungarian word for ’stomach’ and the similar Turkic words for ’internalcavity’, because, in his opinion, this meaning is not present in all the Tur-kic dialects. Instead, J. Budenz connects the Hungarian word for ’stomach’with Ob-Ugric words meaning, variously, ’round’ or ’egg’. The justifica-tion is that these objects all share a round form.3a. According to Á. Vámbéry (1869 : 147) Hun. gyomor ’stomach’ is to beconnected to Cha. tamur ’internal cavity’.3b. According to J. Budenz (1871—1873 : 118) gyomor belongs to Class III. 3c. According to J. Budenz (MUSz 171) this word is to be connected withOst. Énum˙ır ’round’ / Vog. Én∏amr ’egg’. Such meaning association is justi-fied by similar associations to be found in Indo-European languages.3d. gyomor is not listed in UEW; the alleged Vogul and Ostyak corre-spondences are not listed either. This word is nowadays generally con-sidered to be a loan from Turkic (see Ligeti 1986 : 21, 196, etc.).

In example 4 J. Budenz rejects the connection between the Hungarianand Turkic words for ’hair’, on the grounds that the former relates to thehair of the head and the latter to the hair of the body. Instead, he con-nects the Hungarian ’hair of the head’ with Ugric words meaning vari-ously ’hair’, ’fibre’ or ’horsehair’.4a. According to Á. Vámbéry (1869 : 149) Hun. haj ’hair (of head)’ is tobe connected to Cha. kil / Osm. k£ıl, with the same meaning.4b. According to J. Budenz (1871—1873 : 118), haj belongs to Class III; itcannot be connected to the Turkic forms, because their meaning is ’hairof body’.


90

4c. According to J. Budenz (MUSz 129), haj is to be connected to Vot. andZyr. si ’hair (of head), fibre’ / Finn. hiukset ’hair (of head)’. 4d. According to UEW 854, the Hungarian term derives from the Ugricform *kajê, to be connected to Vog. ˛ºoj, kºoj.

Note that Finn. hiukset does not seem to be related to any other wordwithin Uralic (apart from Balto-Finnic, UEW 15). Furthermore, Vot. andZyr. si derive from a totally different root, reconstructed as P-U *sije (*sü-je), which means ’annual ring of a tree, fibre’ and to which Finn. syy, ofthe same meaning, belongs (UEW 443).

It can be seen from these examples that J. Budenz is highly unsys-tematic in how widely he is prepared to interpret the meanings in ques-tion: they may be unacceptably broad to prove one point, and unduly restric-tive to prove another. It is on this kind of basis that the Hungarian/Ugricconnection was in fact founded!

To conclude this list we would like to mention a single example of par-allels which, according to J. Budenz, are only Ugric, but which nowadaysare recognised to be widespread beyond the Uralic area.5. Hun. hÉugy ’urine’ (connected by Á. Vámbéry to Cha. suduk / Osm. s£ıd£ık,etc. (1969 : 153)) / Finn. kusi, etc. J. Budenz (MUSz 120) correctly con-nected these two Hungarian and Finnish etymologies, rejecting Á. Vám-béry’s Turkic/Hungarian connection. However, according to UEW, this term(< P-U *kuÉnÉce ~ *kuÉce) has parallels in all the Tungusic languages, compare,for example, Orok ˛udugu (UEW 210).

4. Statistical analysis

In this section we sh~all perform a statistical analysis of all the 123 Hungar-ian/Ugric parallels under discussion (actually reduced to 119 because ofa few duplets), to verify the rate of agreement between J. Budenz’ vs UEWclassification. In other words, we aim to assess the validity of these Ugriccorrespondences in the light of modern knowledge. For comparison wehave chosen the etymologies and reconstructions proposed by UEW dic-tionary, for the reasons expounded in footnote 8, even though the phono-logical and semantic criteria adopted by UEW in establishing its corre-spondences are rather lax (see again footnote 8).

The complete list of the parallels under examination is given in TableIII at the end of the paper. The table is organized in the following way:the first column lists all the Hungarian items we have dealt with so far,that is a l l the core-Lexicon items proposed by Á. Vámbéry and com-mented by J. Budenz, with their meaning in the second column. J. Budenz’and UEW classification of these items are given in the third and fourthcolumn respectively. The ? symbol in Budenz’ column means that the authorcannot decide whether the good correspondences between Turkic and Hun-garian are indicative of borrowing or of an old, genetic relationship. One ?in UEW means that the classification is considered a bit uncertain by thedictionary, whilst double ?? means that the whole etymology is consid-ered to be very uncertain. In the fifth column, labelled ”Our classification”,we have simplified the original J. Budenz’ and UEW classifications, in orderto have a manageable number of groups that can be analysed statisticallyin a meaningful way. As one can see, the three different categories found


91

in UEW — ”not listed”, ”uncertain etymology (?)”, ”very uncertain etymology(??)”— have all been reduced and marked with one ? only. Similarly, anykind of uncertainty expressed by J. Budenz is simply marked with ?. Notethat, within this fifth column, the format is as follows: ’re-classification ofJ. Budenz / re-classification of UEW’. The last two columns provide thereference of the Hungarian word under discussion(that is, the page num-ber), whereby in the UEW column we have also indicated the specificUralic sub-branch the Hungarian parallel belongs to.

The following graphs illustrate the results of the statistical analysis.

The first graph represents J. Budenz’ classification of the 119 parallels,with the following results:• 25% are considered testimony of an old, genetic relationship betweenHungarian and Turkic; • 10% are still considered good Hungarian/Turkic correspondences, but Bu-denz is unsure whether they are the result of borrowing or real cognates;• 62% are classified as good, pure Hungarian/Ugric correspondences.

As one can see, the overall number of the good Hungarian/Turkic cor-respondences (whether believed to be borrowed or cognates) are 35%,against the 62% of the good Hungarian/Ugric correspondences. This istherefore, according to the author, evidence for a closer relationshipbetween Hungarian and the other Ugric languages than between Hungar-ian and Turkic (but see below).

The second graph represents UEW classification of J. Budenz’ etymologies:• only 28% of J. Budenz correspondences are recognised as good, purelyFinno-Ugric/Uralic etymologies;• 20% are classified as good Finno-Ugric/Uralic etymologies, but they havegood correspondences also in non-Uralic languages, mainly Turkic, Mongo-lian, Tungus, Yukaghir, and, more rarely, in Indo-European languages (U +).• 52% of the etymologies are not recognised as valid anymore.

As one can see, there is little difference between the etymologies clas-sified as purely Uralic (28%) and the etymologies classified as ’U +’ (20%).


92

Finno-U

urkic5%

ic%

20%

Graph 1 Graph 2The core lexicon of the

Uralic paradigm (Budenz)The core lexicon of theUralic paradigm (UEW)

Turkic25%

?Turkic10%

Other3%

Finno-Ugric62%

Uralic+20%

Uralic28%

?52%

The third graph illustrates the degree of agreement between UEW andJ. Budenz’ classification. The calculation has been done on the basis of ourre-classification, that is on the basis of the classification presented in thecolumn labelled ”Our classification” in Table III. The results are as follows:• J. Budenz and UEW agree about the pure Finno-Ugric/Uralic nature ofthe parallels only with regard to 19% of them;• 10% are classified as purely Ugric by J. Budenz, but as having good cor-respondences also in non-Uralic languages (U +) by UEW;• 5% are recognised as good Ugric/Turkic correspondences by J. Budenz,but as pure Uralic etymologies only by UEW;• 8% are recognised as good Ugric/Turkic correspondences by J. Budenz;similarly, UEW recognises that there are correspondences with other non-U languages (U +), in addition to Turkic;• 31% of the parallels classified as Ugric by J. Budenz are classified asuncertain by UEW;• 5% are considered as uncertain etymologies by both sources;• 13% of the parallels are considered as good Turkic/Ugric correspon-dences by J. Budenz, but as uncertain by UEW;• finally, absence of agreement is testified by the other minor classifica-tions: other / ? (3%); ? / U (3%); ? / U + (3%).

As one can see, the number of the etymologies that are classified asbeing purely Ugric/Uralic and for which there is agreement between thetwo sources is very low, only 19%. This number is actually even inferiorto the overall percentage of the good Uralic etymologies which are recog-nised as having good correspondences in non-Uralic languages, at leastaccording to UEW: a total of 21%.

5. Summary and Conclusions

One can summarise the results of the analysis carried out in this paper asfollows.


93

c

10% 13%

Turk

T

?/Ur3

Graph 3

?/Uralic3%

?/Uralic+3%

?/?5%

Finno-Ugric/?31%

The core lexicon of the foundation of the Uralic paradigm:Closest relations according to Budenz/UEW

Finno-Ugric/Uralic19%

Turkic/Uralic+8%

Turkic/Uralic5%

Finno-Ugric/Uralic+10%

Other3% Turkic/?

5%

• A great deal of J. Budenz’ correspondences are incorrect according tomodern knowledge. In addition, the number of errors is higher than onewould expect, even allowing for the historical and factual difficulties encoun-tered in applying the Comparative Method to the languages under dis-cussion (see Marcantonio 2001).• The number of J. Budenz’ as well as UEW good correspondences whichdo not have parallels beyond the Uralic context is not so much higher thanthe number of the good Finno-Ugric/Uralic correspondences for which a cor-respondence with another Asiatic language can be established. In other words,judging from these data only, one could not claim that there is conclusive,compelling evidence in favour of a clear-cut Finno-Ugric/Uralic family —assuming for a moment that the etymologies under investigation are all valid.• As claimed by J. Budenz, there is indeed evidence for a (slightly) closerrelationship between Hungarian and the other Uralic languages, ratherthan between Hungarian and Turkic. However, one must not forget thatthe data are heavily biased toward the Finno-Ugric/Uralic family. Thisbias holds true not only for J. Budenz’ data (as we have seen through someexamples), but also, obviously, for the data reported by UEW and utilisedhere for the purpose of comparison.

Our first conclusion is, therefore, that the existence and uniqueness ofthe Finno-Ugric node was n o t established scientifically and beyond doubtin the last decades of the 19th Century, as widely propagated and believed.

Our second conclusion is that, to borrow D. Sinor’s words, ”... Uralic, Altaic,and Uralo-Altaic comparative linguistics should shake themselves free fromsimplistic — black and white, yes and no — solutions” (1988 : 739) and that,therefore, the traditional approach to the Uralic and Altaic studies clearlydemands a much needed revision.

Table IJ. Budenz’ Class I ”correct correspondences”

(between Hungarian and Turkic)

Table IaThe correspondences indicative of genetic relations

Hungarian Reference UEWclassification

Éangy ’sister / daughter-in-law, MUSz 746 U +, 10aunt’

anya ’mother’ MUSz 745 U +, 10 (?)apa ’father’ MUSz 746 F-U, 14atya ’father-in-law’ MUSz 760 n.l.öcs ’younger brother’ MUSz 846 F-U +, 70agg ’old man’ MUSz 717 F-U, 448gyerek ’child’ MUSz 783 n.l.eméÓo/emse ’breast, mother / sow’ MUSz 783 U +, 74hÉat ’back’ NyK 83 U, 225kar ’arm’ MUSz 5 n.l.far(k)/ ’tail, back-part, bottom / MUSz 495 Ug., 407 (??)farkas wolf’bor-ulni ’to get cloudy’ NyK 78 n.l.csalit ’scrub, bush’ NyK 79 n.l.


94

domb ’hill’ MUSz 256 Ug., 896felhéÓo ’cloud’ MUSz 511 F-U, 381fÉeszek ’nest’ MUSz 520 U, 357kova/kéÓo ’stone’ MUSz 40 F-U +, 163sÉar ’mud’ NyK 87 n.l.szaj/zaj ’thin ice’ MUSz 266 F-U, 29szÉal ’thread / stalk’ MUSz 270 F-U +, 460alud-ni/ ’to sleep’ MUSz 270 F-U +, 334al-szikem-ni ’to suck’ MUSz 781 U +, 82fÉuj-ni ’to blow’ MUSz 540 U +, 411gyéÓul-ni/ ’to assemble’ MUSz 165-166 n.l.gyéÓujte-nihat-ni ’to act’ NyK 83 U, 130jÉar-ni ’to go, walk’ MUSz 140 F-U, 102jön-ni, jöv-ök ’to come’ MUSz 153 Ug., 109men-ni, megyek ’to go’ MUSz 611-612 U +, 272vol(t), val(a), ’to be, become’ MUSz 592 F-U +, 580vagyal, alatt ’under’ MUSz 728 U +, 6alsÉo ’lower, below’ MUSz 728 n.l.eléÓo/eléÓott ’pre / ahead, in front’ MUSz 778 F-U +, 71-72éert, éerett ’for, because of’ MUSz 794 n.l.csel-ekedni ’to do, act’ MUSz 363 n.l.

Table IbThe correspondences indicative of borrowing

Hungarian Reference UEWclassification

béÓor ’skin’ NyK 79 U +, 374csecs, cséecs ’breast, udder, nipple’ MUSz 374 n.l.daru ’crane’ Nyk 80 F-U +, 513hajÉo ’boat’ MUSz 72-73 U +, 169-170(?)

hangya ’ant’ NyK 83 F-U, 209hÉod ’beaver’ NyK 83 Ug +., 858köd ’fog’ NyK 86 U, 158serte/sertÉes ’bristle / pork, pig’ NyK 87 n.l.

Table IIJ. Budenz’ Class II and III Ugric correspondences

recognised as good Finno-Ugric/Uralic correspondences and for which there are no parallels in other language families (according to UEW)

Hungarian Budenz UEWclassification

rokon ’relative, related’ MUSz 668 F-U, 418epe ’bile’ MUSz 791 F-U, 435fül ’ear’ MUSz 548 F-U, 370haj ’hair’ (of the head) MUSz 129 Ugric, 854nyÉal ’saliva’ MUSz 406-407 F-U, 322


95

Hungarian Budenz UEWclassification

széıv ’heart’ MUSz 300 U, 477veléÓo ’marrow’ MUSz 571 F-U, 572¥eb ’dog’ MUSz 764 Ugric, 836harkÉaly ’woodpecker’ NyK 83 Ugric, 855csillag ’star’ MUSz 367-368 F-U, 46éeg ’sky, heavens’ MUSz 511 F-U, 435féÓu ’grass’ MUSz 545 Ugric, 878gyökéer ’root’ MUSz 163 Ugric, 852hagyma ’onion, bulb’ NyK 101 F-U, 164éÓosz ’autumn’ MUSz 859 F-U, 443tavasz ’spring’ MUSz 187 F-U, 532éeg ’to burn’ MUSz 766 F-U, 26foly-ik ’to flow, run’ MUSz 530 Ugric, 881 fut ’to run’ MUSz 544-545 U, 402iv- (i-, isz) ’to drink’ MUSz 826-827 F-U, 103nyal ’to lick’ MUSz 407- 408 U, 321Éeh, Éehes ’hungry’ MUSz 771 Ug., 851gyalog ’by foot’ MUSz 160 F-U, 88-89

Table IIIAll J. Budenz’ (Class I—III) correspondences

and their classification according to UEW

Hungarian Budenz UEW Our Ref. Ref.classification Budenz UEW

daru ’crane’ ?Turkic ?Uralic&A ?Turkic/? NyK10,80 F-U 513(borrowed)

serte, ’bristle, ?Turkic Not listed ? Turkic/? MUSz 312sertées pig/pork’ (borrowed)Éert, ’for, pro’ ?Turkic Not listed ? Turkic/? MUSz 794Éerett (genetic)fark/ ’tail/ ?Turkic Not listed ? Turkic/? NyK 10, 81farkas wolf’ (genetic)csalit ’thicket, ?Turkic Not listed ? Turkic/? NyK 10, 79

copse’ (genetic)sÉar ’mud’ ?Turkic Not listed ?Turkic/? NyK 10, 87

(genetic)fÉeszek ’nest’ ?Turkic Uralic ?Turkic/ MUSz 520 U 375

(genetic) Uraliczaj, ’thin ice’ ?Turkic Uralic ?Turkic/ MUSz 266 F-U 29szaj (genetic) Uralichat ’to act, ?Turkic Uralic ? Turkic/ NyK 10, 83 U 130

be effective’ (genetic) UralichÉod ’beaver’ ?Turkic Uralic & A ? NyK 10, 83 Ug. 858

(borrowed) Turkic/Uralic+nyÉar ’summer’ ?Turkic Uralic & Yuk ? NyK 10, 106 U 324

(borrowed) Turkic/Uralic+men-ni, ’to go, ?Turkic Uralic & Yuk ? MUSz 611 U 272megy travel’ (genetic) Turkic/Uralic+orr ’nose’ Finno-Ugric ?Hungarian Finno-Ugric/? MUSz 824 F-U 571


96


fÉerfi ’man’ Finno-Ugric ?Uralic & A Finno-Ugric/? MUSz 517 F-U 84hÉaj ’fat/ Finno-Ugric ?Uralic & A Finno-Ugric/? MUSz 70 F-U 195

lard’Éur ’gentle- Finno-Ugric ?? Finno-Ugric/? MUSz 690 F-U 545

man’béel ’intestines’ Finno-Ugric ?? Finno-Ugric/? MUSz 453 F-U 364csont ’bone’ Finno-Ugric ?? Finno-Ugric/? NyK 10, 79 F-U 45nyak ’neck’ Finno-Ugric ?? Finno-Ugric/? MUSz 405 U 328zÉap/ ’rotten/ Finno-Ugric ?? Finno-Ugric/? MUSz 323 Ug 885zÉap(fog) molar

tooth’far ’bottom’ Finno-Ugric ?? Finno-Ugric/? MUSz 495 Ug. 407nyéÓu ’worm’ Finno-Ugric ?? Finno-Ugric/? MUSz 428 Ug. 320toj-ik ’to lay Finno-Ugric ?? Finno-Ugric/? MUSz 222 F-U

eggs’ 528–529falu ’village’ Finno-Ugric ?? Finno-Ugric/? MUSz 492 Ug. 351köröm ’(finger-) Finno-Ugric ??Hungarian Finno-Ugric/? MUSz 44 F-P 677

nail’toll ’feather’ Finno-Ugric ?Uralic & A Finno-Ugric/? MUSz 225 U 535bÉacsi ’uncle’ Finno-Ugric Not listed Finno-Ugric/? NyK 10, 94hÉug ’younger Finno-Ugric Not listed Finno-Ugric/? NyK 10,101

sister’szajkÉo ’jay/ Finno-Ugric Not listed Finno-Ugric/? NyK 10, 109

parrot’özön ’stream, Finno-Ugric Not listed Finno-Ugric/? NyK 10, 108

flood’kölyök ’young Finno-Ugric Not listed Finno-Ugric/? MUSz 42

of animal’legéeny ’young Finno-Ugric Not listed Finno-Ugric/? MUSz 690

man’boka ’ankle, Finno-Ugric Not listed Finno-Ugric/? MUSz 460

heels’gyomor ’stomach’ Finno-Ugric Not listed Finno-Ugric/? MUSz 171koponya ’skull, Finno-Ugric Not listed Finno-Ugric/? NyK 10, 120

head’köldök ’navel’ Finno-Ugric Not listed Finno-Ugric/? MUSz 41talp ’sole’ Finno-Ugric Not listed Finno-Ugric/? MUSz 175téerd ’knee’ Finno-Ugric Not listed Finno-Ugric/? NyK 10, 125gebe ’nag, Finno-Ugric Not listed Finno-Ugric/? NyK 10, 99

weakperson’

kutya ’dog’ Finno-Ugric Not listed Finno-Ugric/? MUSz 52szöcske ’grass- Finno-Ugric Not listed Finno-Ugric/? NyK 10, 110

hopper’beng ’seed, Finno-Ugric Not listed Finno-Ugric/? NyK 10, 115[sic] bean

(of somefruit)’

benge ’bud’ Finno-Ugric Not listed Finno-Ugric/? NyK 10, 95[sic]


972 Linguistica Uralica 2 2001


déer ’(hoar-) Finno-Ugric Not listed Finno-Ugric/? MUSz 248frost’

deréÓu ’clear sky/ Finno-Ugric Not listed Finno-Ugric/? MUSz 249weather’

korom ’soot’ Finno-Ugric Not listed Finno-Ugric/? NyK 10, 104verem/ ’hole/ Finno-Ugric Not listed Finno-Ugric/? MUSz 874üres empty’gyÉul ’to catch Finno-Ugric Not listed Finno-Ugric/? MUSz 190

fire’dééel ’south, Finno-Ugric Not listed Finno-Ugric/? NyK 10, 97

noon’gyanÉant ’as, Finno-Ugric Not listed Finno-Ugric/? NyK 10, 100

by way of’rokon ’relative, Finno-Ugric Uralic Finno-Ugric/ MUSz 668 F-U 418

related’ Uralicepe ’bile’ Finno-Ugric Uralic Finno-Ugric/ MUSz 791 F-U 435

Uralicfül ’ear’ Finno-Ugric Uralic Finno-Ugric/ MUSz 548 F-U 370

Uralichaj ’hair Finno-Ugric Uralic Finno-Ugric/ MUSz 129 Ug. 854

(of the Uralichead)’

nyÉal ’saliva’ Finno-Ugric Uralic Finno-Ugric/ MUSz F-U 322Uralic 406–407

széıv ’heart’ Finno-Ugric Uralic Finno-Ugric/ MUSz 300 U 477Uralic

veléÓo ’marrow’ Finno-Ugric Uralic Finno-Ugric/ MUSz 571 F-U 572Uralic

¥eb ’dog’ Finno-Ugric Uralic Finno-Ugric/ MUSz 764 Ug. 836Uralic

harkÉaly ’wood- Finno-Ugric Uralic Finno-Ugric/ NyK 10, 83 Ug. 855pecker’ Uralic

csillag ’star’ Finno-Ugric Uralic Finno-Ugric/ MUSz F-U 46Uralic 367–368

éeg ’sky, Finno-Ugric Uralic Finno-Ugric/ MUSz 511 F-U 435heavens’ Uralic

féÓu ’grass’ Finno-Ugric Uralic Finno-Ugric/ MUSz 545 Ug. 878Uralic

gyökéer ’root’ Finno-Ugric Uralic Finno-Ugric/ MUSz 163 Ug. 852Uralic

hagyma ’onion, Finno-Ugric Uralic Finno-Ugric/ NyK 10, 101 F-U 164bulb’ Uralic

éÓosz ’autumn’ Finno-Ugric Uralic Finno-Ugric/ MUSz 859 F-U 443Uralic

tavasz ’spring’ Finno-Ugric Uralic Finno-Ugric/ MUSz 187 F-U 532Uralic

éeg-ni ’to burn’ Finno-Ugric Uralic Finno-Ugric/ MUSz 766 F-U 26Uralic

foly-ik ’to flow, Finno-Ugric Uralic Finno-Ugric/ MUSz 530 Ug. 881run’ Uralic


98

fut ’to run’ Finno-Ugric Uralic Finno-Ugric/ MUSz U 402Uralic 544–545

iv- ’to drink’ Finno-Ugric Uralic Finno-Ugric/ MUSz F-U 103(i-, isz) Uralic 826–827nyal ’to lick’ Finno-Ugric Uralic Finno-Ugric/ MUSz U 321

Uralic 407– 408Éeh, ’hungry’ Finno-Ugric Uralic Finno-Ugric/ MUSz 771 Ug. 851Éehes Uralicgyalog ’by foot’ Finno-Ugric Uralic Finno-Ugric/ MUSz 160 F-U 89

UralicÉut ’way, road’ Finno-Ugric Uralic Finno-Ugric/ MUSz 869 U 546

UralichÉugy ’urine’ Finno-Ugric Uralic & A Finno-Ugric/ MUSz 120 U 210

Uralic+mÉaj ’liver’ Finno-Ugric Uralic & A Finno-Ugric/ MUSz 599 U 264

Uralic+hab ’surf, foam’ Finno-Ugric Uralic & A Finno-Ugric/ NyK 10, 101 U 203

Uralic+éın ’tendon, Finno-Ugric Uralic & I-E Finno-Ugric/ MUSz 817 U 441

sinew’ Uralic+fÉel ’half’ Finno-Ugric Uralic & I-E Finno-Ugric/ MUSz 506 U 362

Uralic+éÓos ’ancestor’ Finno-Ugric Uralic & Yuk Finno-Ugric/ MUSz 857 U 78

Uralic+hÉo ’snow’ Finno-Ugric Uralic & Yuk Finno-Ugric/ MUSz 136 U 204

Uralic+en-ni, ’to eat’ Finno-Ugric Uralic & Yuk Finno-Ugric/ MUSz 802 F-U 440esz-ik Uralic+len-ni/ ’to be, Finno-Ugric Uralic & Yuk Finno-Ugric/ MUSz 698 F-U 243lesz become/ Uralic+

will be’Éaltal/ ’across/ Finno-Ugric Uralic & A Finno-Ugric/ MUSz 740 U 332Éat through, Uralic+

by means of,’hegy ’mountain, Finno-Ugric Uralic & Yuk Finno-Ugric/ MUSz 96 F-U 110

hill, top’ Uralic+hal ’fish’ Finno-Ugric Wanderwort Finno-Ugric/ MUSz 77 U 119

Uralic+keczel ’puppy’ Rumanian Not listed Other/? NyK 10, 120gerinc ’spine, Slavic Not listed Other/? NyK10, 117

backbone’csorda ’herd’ Slavic Not listed Other/? NyK 10, 116hajÉo ’boat’ Turkic ??Hungarian Turkic/? MUSz 73 U 169

(borrowed)béÓor ’skin’ Turkic ?Uralic & A Turkic/? NyK 10, 79 U 374

(borrowed)csecs ’breast, Turkic Not listed Turkic/? MUSz 374

udder’ (borrowed)jÉar ’to go, Turkic ? Hungarian Turkic/? MUSz 140 F-U 102

to function’ (genetic)öcs ’younger Turkic ? Uralic & A Turkic/? MUSz 846 F-U 70

brother’ (genetic)


992*


domb ’hill’ Turkic ?? Turkic/? MUSz 256 Ug. 896(genetic)

jön, ’to come’ Turkic ?? Turkic/? MUSz 153 Ug. 109jövök (genetic)eléÓo, ’fore (part)’ Turkic ??Uralic & A Turkic/? MUSz 778 F-U 72eléÓott (genetic)

anya ’mother’ Turkic ?Hungarian Turkic/? MUSz 745 U 10(genetic)

atya ’father’ Turkic Not listed Turkic/? MUSz 760(genetic)

gyerek ’child’ Turkic Not listed Turkic/? MUSz 171(genetic)

kar ’arm’ Turkic Not listed Turkic/? MUSz 5(genetic)

borul ’to become’ Turkic Not listed Turkic/? NyK 10, 78cloudy (genetic)

csel- ’to do, act’ Turkic Not listed Turkic/? MUSz 363(eksz-ik) (genetic)gyéÓujt- ’to assemble, Turkic Not listed Turkic/? NyK 10, 82

gather’ (genetic)köd ’fog’ Turkic Uralic Turkic/ NyK 10, 86 U 158

(borrowed) Uralichangya ’ant’ Turkic Uralic Turkic/ NyK 10, 83 F-U 209

(borrowed) Uralicapa ’father’ Turkic Uralic Turkic/ MUSz 746 F-U 14

(genetic) Uralicagg ’old man’ Turkic Uralic Turkic/ MUSz 717 F-U 448

(genetic) UralichÉat ’back’ Turkic Uralic Turkic/ NyK, 10,83 U 225

(genetic) UralicfelhéÓo ’cloud’ Turkic Uralic Turkic/ MUSz 511 F-U 381

(genetic) UralicÉangy ’sister-in-law’ Turkic Uralic & A Turkic/ MUSz 746 U 10

(genetic) Uralic+em-ni, ’to suck’ Turkic Uralic & A Turkic/ MUSz 781 U 82emik (genetic) Uralic+fÉuj ’to blow’ Turkic Uralic & A Turkic/ MUSz 540 U 411

(genetic) Uralic+vol-, ’to be, Turkic Uralic & A Turkic/ MUSz 592 F-U 580vala etc. become’ (genetic) Uralic+al-, ’[space] Turkic Uralic & A Turkic/ MUSz 728 U 6alatt under’ (genetic) Uralic+ékÓéÓo ’stone’ Turkic Uralic & Yuk Turkic/ MUSz 40 F-U 163

(genetic) Uralic+szÉal ’thread, Turkic Uralic & Yuk Turkic/ MUSz 270 F-U 460

stalk’ (genetic) Uralic+eméÓo [sic]/ ’breast/ Turkic Uralic & A Turkic/ MUSz 783 U 74emse sow’ (genetic) Uralic+alud-, ’to sleep’ Turkic Uralic & A Turkic/ MUSz 737 F-U 334alsz-ik (genetic) Uralic+


100

Languages and language-groups abbreviations

A — Altaic; Cha. — Chagatay; Cher. — Cheremis; Est. — Estonian; Finn. — Finnish;F-U — Finno-Ugric; Ger. — German; Hun. — Hungarian; I-E — Indo-European;Lat. — Latin; Mord. — Mordvin; MordM — Mordvin (Moksha); Old Ind. — OldIndian; Osm. — Osmanli (Turkish); Ost. — Ostyak; P-U — Proto Uralic; P-F-U —Proto-Finno-Ugric; Tat. — Tatar; Turk. — Turkic (languages); U — Uralic; Ug. —Ugric (= Ugric languages according to current meaning: Vogul, Ostyak, Hungar-ian); Vog. — Vogul; Vot. — Votyak; Zyr. — Zyrian; Yuk. — Yukaghir.

B I B L I O G R P H Y

A b o n d o l o, D. 1987, Hungarian. — The Major Languages of Eastern Europe,London, 185—200.

B u d e n z, J. 1869, A magyar és finn-ugor nyelvekbeli szóegyezések. — NyK 6,374—478.

—— 1871—1873, Jelentés Vámbéry Á. magyar-török szóegyezéseiré Óol. — NyK 10,67—135.

—— 1873—1881, Magyar-ugor összehasonlító szótár, Budapest (= MUSz).—— 1878/1879, Ueber die Verzweigung der ugrischen Sprachen. — Festschrift

zur Feier seines fünfzigjährigen Doctorjubiläums am 24. October 1878Herrn Professor Theodor Benfey gewidmet, Göttingen (Beiträge zur Kundeder indogermanischen Sprachen 4. Separat), 192—258.

D o n n e r, O. 1879/1880, Die gegenseitige Verwandtschaft der finnisch-ugrischenSprachen. — Acta Societatis Scientiarum Fennicae 11, Helsinki, 406—567.

F o x, A. 1995, Linguistic Reconstruction. An Introduction to the Theory andMethod, Oxford.

G e o r g, S., M i c h a l o v e, P. A., M a n a s t e r R a m e r, A., S i d w e l l, P. J.1998, Telling General Linguists about Altaic. — Journal of Linguistics 35,65—98.

H e l i m s k i, E. 1984, Problems of Phonological Reconstruction in Modern UralicLinguistics. — SFU XX, 241—257.

H u n f a l v y, P. 1883, Ugor vagy török-tatár eredet é Óu-e a magyar nemzet. —Értekezések a nyelv és széptudományok köréb é Óol 1/11, 1—39

K á l m á n, B. 1988, The History of the Ob-Ugric Languages. — The Uralic Lan-guages. Description, History and Foreign Influences. Handbook of UralicStudies I, Leiden—New York, 394—412.

K u l o n e n, U.-M. 1996, Sanojen alkuperä ja sen selittäminen. Etymologista lek-sikografiaa, Helsinki (Suomi 181).

L i g e t i, L. 1986, A magyar nyelv török kapcsolatai a honfoglalás el é Óott és azÁrpád-korban, Budapest.

M a r c a n t o n i o, A. 2001 (forthcoming), The Uralic Language Family. Facts,Myths and Statistics, Oxford (Transaction of the Philological Society).

P u s z t a y, J. 1977, Az ”ugor-török háború” után. Fejezetek a magyar nyelvha-sonlítás történetéb é Óol. Budapest.

R ä s ä n e n, M. 1963/1964, Uralaltailaisesta kielisukulaisuudesta. — Esitelmät japöytäkirjat, Helsinki (Suomalainen Tiedeakatemia), 180—189.

R ó n a - T a s, A. 1978, Julius Németh: Life and Work, Budapest (Acta OrientaliaAcademiae Scientiarum Hungaricae 32), 261—236.

—— 1988, Turkic Influence on the Uralic Languages. — The Uralic Languages.Description, History and Foreign Influences. Handbook of Uralic Studies I,Leiden—New York, 742—780.

R u h l e n, M. 1987, A Guide to the World’s Languages I. Classification, Lon-don—Melbourne.

S a l v a g n i, M. 1999, La ”guerra ugro-turca” e l’affermarsi della teoria ugro-finnica. Tesi di Laurea, Roma (Università di Roma).

S a m m a l l a h t i, P. 1988, Historical Phonology of the Uralic Languages (withSpecial Reference to Samoyed, Ugric and Permic). — The Uralic Languages.Description, History and Foreign Influences. Handbook of Uralic Studies I,Leiden—New York, 478—554.


101

S i n o r, D. 1988, The Problem of the Ural-Altaic Relationship. — The UralicLanguages. Description, History and Foreign Influences. Handbook of UralicStudies I, Leiden—New York, 706—741.

S z i n n y e i, J. 1896—1927, Magyar nyelv hasonlítás, Budapest.Suomen sanojen alkuperä. Etymologinen sanakirja 1. A—K, Helsinki 1992 (SKST

556) (= SSA).S u t r o p, U. 2000, The Forest of Finno-Ugric Languages. — The Roots of Peo-

ples and Languages of Northern Eurasia II and III, Tartu (FU 23), 165—197.

V á m b é r y, Á. 1869, Magyar és török-tatár szóegyezések. — NyK 8, 109—189.—— 1882, A magyarok eredete; ethnologiai tanulm Éany, Budapest.

ANDWELA MARKANTONIO (Rim), PIR≤O NUMMENAHO (Neapolx),MIKELA SALVAN≤I (Rim)

«UGORSKO-TŒRKSKAQ BOR≤BA»: KRITIÄESKI| OBZOR

V poslednie desqtiletiq XIX veka v uralistike s pomoYxœ nauänyh metodovanaliza bylo ustanovleno suYestvovanie unikalxnoj finno-ugorskoj qzykovojsemxi. Åta teoriq v osnovnom bazirovalasx, kak predpolagalosx, na sravnitelxnyhissledovaniqh nemeckogo uäenogo |. Budenca, rabotavöego v Vengrii v konceXIX veka.

Prinqto säitatx, äto |. Budenc ustanovil unikalxnostx finno-ugorskoj semxii na ego vyvodah osnovany mnogie issledovaniq. Odnako do sih por v literaturene obsuwdalisx sami originalxnye trudy |. Budenca. Predlagaemaq statxq po-svqYena ih kritiäeskomu peresmotru.

Dve naibolee vawnye problemy, na kotoryh hotelosx by ostanovitxsq: vo-per-vyh, |. Budenc, nesmotrq na svoi utverwdeniq, sam upomqnutym sravnitelxnymmetodom ne polxzovalsq v svoih trudah; vo-vtoryh, ego vyvody ob unikalxnostifinno-ugorskoj semxi ne sovpadaœt s predpoloweniqmi sovremennyh issledo-vatelej. Ustanovlennye |. Budencem mnogoäislennye turecko-vengerskie ana-logii svidetelxstvuœt skoree o genetiäeskih svqzqh vengerskogo i tureckogoqzykov!


102

the ”ugric-turkic battle”: a critical review … · the ”ugric-turkic battle”: a critical...

Documents